The BC development world went from "everything was on the CD" to a wall of technologies we're all expected to know. And now, agentic AI tools are adding another layer on top.

Now, we're in another big rewrite of what it means to be a "BC Developer". Unlike past talks where I've covered how much the toolings and day-to-day expectations of what we must know to deliver code, this time we have a fundamental shift in our processes and even way we think about getting work done. That's a role change.

In this session, I'll walk through what I've learned from months of daily agentic development and from working with teams of senior BC developers going through this transition - some willingly, some reluctantly, and some under pressure from management.

We'll look at the five responsibilities that are replacing "BC Developer" as a job description: Requirements Architect, Quality Gatekeeper, Context Engineer, Pattern Guardian, and Fleet Manager. For each one, I'll show what the work actually looks like now - with live demos of the prompts, instructions, validation workflows, and design conversations that make up a modern BC developer's day.

And finally, some practical guidance on how to navigate the times ahead. Not everyone in this room is in the same situation — whether you're being pushed into this, pulling your org along behind you, or just trying to figure out where you fit — you'll leave with a concrete action plan for Monday morning.

GitHub Copilot is awesome. It helps you write code.
…but you already knew that.

What about everything around the code?

Requirements. Test plans. Code reviews. Documentation. Design notes.
All the parts of the development workflow that usually get rushed, postponed, or done “just enough.”

In this hands-on workshop, we’ll explore how to use GitHub Copilot for exactly those non-coding tasks by working with custom agents, sub-agents, MCPs, and stored prompts.

We’ll cover:

• How to create custom agents for tasks like requirement analysis, test planning, and documentation
• How these compare to community-built agents
• Where sub-agents fit into more complex workflows
• What MCPs are, and which ones work best for non-coding use cases
• How to prepare prompts for reliable one-time execution

You’ll leave with ready-to-use prompts and agent configurations that you can share with your team and apply to your own projects the very next day.

You’ve heard a lot about using AI to write code.
Join this workshop to learn how to use it for everything else.

Audience:

• Development Managers
• Developers
• Testers
• Product Owners

Workshop date: Tue 9 or Wed 10 June (This is a one-day workshop)

The workshop on 10 June is sold out!